Probing the Hydrogen-Bonded Water Network at the Active Site of a

May 22, 2015 - bonding network present at the active site a crucial part of the catalytic mechanism. In this paper, we present the spectroscopic study...
0 downloads 8 Views 3MB Size
Download PDF
Article pubs.acs.org/JPCA

Probing the Hydrogen-Bonded Water Network at the Active Site of a Water Oxidation Catalyst: [Ru(bpy)(tpy)(H2O)]2+·(H2O)0−4 Erin M. Duffy, Brett M. Marsh, and Etienne Garand* Department of Chemistry, University of Wisconsin, 1101 University Avenue, Madison, Wisconsin 53706, United States S Supporting Information *

ABSTRACT: The infrared spectra of gas-phase mass-selected [Ru(bpy)(tpy)(H2O)]2+·(H2O)0−4 clusters (bpy = 2,2′-bipyridine; tpy = 2,2′:6,2″-terpyridine) in the OH stretching region are reported. These species are formed by bringing the homogeneous water oxidation catalyst [Ru(bpy)(tpy)(H2O]2+ from solution into the gas phase via electrospray ionization (ESI) and reconstructing the water network at the active site by condensing additional water onto the complex in a cryogenic ion trap. Infrared predissociation spectroscopy is used to probe the structure of these clusters via their distinctive OH stretch frequencies, which are sensitive to the shape and strength of the local hydrogen-bonding network. The analysis of the spectra, aided by electronic structure calculations, highlights the formation of strong hydrogen bonds between the aqua ligand and the solvating water molecules in the first solvation shell. These interactions are found to propagate through the subsequent solvation shells and lead to the stabilization of asymmetric solvation motifs. Electronic structure calculations show that these strong hydrogen bonds are promoted by charge transfer from the H atom of the aqua ligand to the Ru−OH2 bond.

■

reaction pathways.4,8,17−22 In particular, Berlinguette and coworkers,17 using rapid mixing electrospray ionization and mass spectrometry, identified masses corresponding to the [Ru− OH]2+, [RuO]2+, and [Ru−O2]2+ reaction complexes, pointing to a catalytic cycle that involves stepwise removal of e−/H+ pairs. For example, in the first step of the oxidation reaction, the aqua ligand on the catalyst is activated, and with e−/H+ removal, the intermediate complex [Ru−OH]2+ is formed. A powerful tool for studying the solvation structure of ionic complexes is infrared predissociation spectroscopy of massselected clusters.23−30 By acquiring the vibrational spectra of isolated and mass-selected ionic species, detailed information concerning structure and bonding can be revealed, providing characteristic trends as a function of solvation.24,26,28−32 In this paper, we use this approach to examine the interactions between the solvent water molecules and the active site of the [Ru(H2O)]2+ catalyst. We focus on the OH stretching region where the sensitivity of the vibrational frequency to local solvation environment serves as an exquisite probe of the hydrogen-bond network. Determination of the cluster geometries is facilitated by comparison with density functional theory (DFT) calculations. Together, the results provide new insight into the interactions between the aqua ligand and the surrounding solvent water molecules.

INTRODUCTION An efficient artificial photosynthesis process needs to overcome the thermodynamic bottleneck of the water oxidation reaction in which O2 is generated from H2O. Mediating this reaction through catalysis has been of significant interest for several decades.1,2 Excitingly, recent developments have shown examples of homogeneous single-metal-center complexes capable of catalyzing water oxidation.3 The structural simplicity afforded by these single-site catalysts makes them significantly more accessible synthetically and easier to study computationally.4−6 This discovery transformed the field and led to development of numerous mononuclear water oxidation catalysts (WOCs).7−10 However, despite several detailed studies, the precise mechanism by which these WOCs operate is still unclear.11,12 One critical unknown aspect is the extent of the role that proton-coupled electron transfer (PCET) plays in the catalytic cycle. PCET can minimize charge buildup at the reaction site13 and therefore avoid the formation of high-energy intermediate complexes.14 However, this process must be assisted by nearby solvent molecules,15,16 making the hydrogenbonding network present at the active site a crucial part of the catalytic mechanism. In this paper, we present the spectroscopic study of the stepwise microsolvation and hydrogen-bonding network around the active site of the well-known [Ru(bpy)(tpy)(H2O)]2+ ([Ru(H2O)]2+) water oxidation catalyst.3,4,8,17 This species, along with its derivatives, has been the subject of many studies targeted at characterizing its catalytic abilities and © XXXX American Chemical Society

Received: May 19, 2015

A

DOI: 10.1021/acs.jpca.5b04778 J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

■

■

EXPERIMENTAL DETAILS

Article

RESULTS AND ANALYSIS The infrared predissociation spectra of the [Ru(H2O)]2+· (H2O)n clusters in the 2500−3800 cm−1 region are presented in Figure 1. The spectra display features in two distinct regions.

The [Ru(H2O)](ClO4)2 catalyst was synthesized and isolated according to published procedures.4,33 RuCl3·3H2O was purchased from Pressure Chemical Co., and all other reagents were obtained from Sigma-Aldrich. The final product identity was confirmed by 1H NMR. The vibrational spectra of the [Ru(H2O)]2+·(H2O)n (n = 0− 4) clusters were obtained using a home-built cryogenic ion photofragmentation mass spectrometer,34 described in detail previously.35 The full description of the experimental details are provided in the Supporting Information. Briefly, the [Ru(H2O)]2+ ions were generated by electrospray ionization (ESI) and collected in a temperature-controlled 3D quadrupole ion trap. To create the microsolvated ions, a small amount of water was seeded in the trap’s helium buffer gas. Formation of [Ru(H2O)]2+(H2O)1−4 clusters was observed when the trap temperature was held at 165 or 185 K, with the lower temperature favoring the formation of [Ru(H2O)]2+(H2O)4 clusters. To obtain the IR spectrum, the mass-selected ion packet of interest was intersected with the output of a tunable infrared OPO/OPA laser system (Laservision). When the IR photon energy is resonant with a vibrational transition, the absorption of a single photon is sufficient to induce the photofragmentation of the ions, and the fragment intensity as a function of photon wavelength yields the IR spectra for the parent cluster. For the [Ru(H2O)]2+ ion, the higher binding energy of the aqua ligand made it impossible to observe singlephoton fragmentation. Instead, we utilized the “D2-tagging” approach in which the ion trap temperature was held at 10 K with 10% D2 added to the buffer gas. Under these conditions, weakly bound D2 molecules condensed onto the ions, forming [Ru(H2O)]2+·(D2)n clusters, and served as the messenger for predissociation spectroscopy. To aid the assignment of the experimental infrared spectra, density functional theory (DFT) calculations were performed using the Gaussian 09 software package.36 Full computational details can be found in the Supporting Information. Three different DFT functionals were used to optimize the geometry, compute the harmonic vibrational spectrum, and calculate the relative energy of various isomers: cam-B3LYP, cam-B3LYP with Grimme’s D3 dispersion scheme, and ωB97XD. For the ruthenium atom, we used the Stuttgart/Dresden effective core potential for 28 core electrons (MWB28) and the Dunning/ Huzinaga full double-ζ (D95) basis set for the remaining electrons. The 6-311+G(d,p) basis set was used for all other atoms. The reported relative energies are corrected for zeropoint energy (harmonic values). The three different functionals yielded slightly different isomer energetic ordering but similar harmonic IR spectra. Additional uncertainty is also involved in the thermal correction which neglects internal rotation of the water molecules. Given the uncertainty in relative energy, we use the calculated IR spectra as the primary tools for assigning the experimental spectra, and we present the cam-B3LYP results here because it provided the best agreement between the calculated and the experimental spectra. The results obtained using the other methods are included in the Supporting Information. To facilitate comparison of experimental and harmonic spectra, the calculated spectra are Gaussian broadened (with area corresponding to the calculated intensity) and the frequencies are scaled by 0.9476.

Figure 1. Infrared predissociation spectra of [Ru(H2O)]2+(H2O)n, n = 0−4, in the 2500−3800 cm−1 region.

For all species, several sharp peaks are present in the 3550− 3750 cm−1 region, which is characteristic of the stretching vibrations corresponding to free OH (i.e., not hydrogen-bond donating). With the addition of the first solvent water molecule, broad and intense features representative of the hydrogenbonded (H-bonded) OH stretch appear below 3550 cm−1. This red shift in the frequency of the OH mode, which is roughly dependent on the square of the H-bond interaction strength,37−39 provides clear experimental evidence that the solvation network is formed around the aqua ligand in these clusters. The spectrum of [Ru(H 2 O)] 2+ is relatively simple, containing only two sharp features at 3576 and 3654 cm−1. The spectrum of [Ru(H2O)]2+·(H2O) contains three sharp peaks in the free OH region at 3631, 3660, and 3712 cm−1 as well as a broad and intense feature at 3200 cm−1. With the addition of the second solvent water molecule, the spectrum of [Ru(H2O)]2+·(H2O)2 displays an even more red shifted OH stretch at 2930 cm−1 and a broad and intense feature at 3355 cm−1 with a shoulder at 3290 cm−1. The spectrum also contains only two well-resolved peaks in the free OH region at 3633 and 3717 cm−1. The spectrum of [Ru(H2O)]2+·(H2O)3 shows two similarly intense and broad features at 3170 and 3360 cm−1, B

DOI: 10.1021/acs.jpca.5b04778 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

2930 cm−1. The experimental spectrum only shows a very faint hint of these weak transitions. For the [Ru(H2O)]2+·(H2O)1 species, only one stable structure was found in the calculations, in which the first solvent water is H-bonded directly to the aqua ligand. The structure and calculated spectrum are shown in Figure 2. The calculated vibrational spectrum is in fairly good agreement with the experimental spectrum, allowing for a straightforward assignment. The intense and broad feature at 3200 cm−1 is assigned to the H-bond-donating OH of the aqua ligand (LD), while the peak at 3660 cm−1 is assigned to the free OH of the aqua ligand (LF). The two remaining peaks at 3631 and 3712 cm−1 correspond to the symmetric and antisymmetric OH stretches of the H-bond-accepting solvent water (S1A). The major discrepancy between the experimental and the calculated spectra is the presence of a shoulder feature at 3300 cm−1. Because no other isomer could be found, we tentatively attribute this feature to anharmonic effects involving either the overtone of the H2O bending mode or simply broadening due to the thermal motion of the water molecules.40 The weaker and broader appearance of the S1A antisymmetric stretch is discussed in further detail in the Supporting Information. The addition of the second solvent water yields two possible configurations, shown in Figure 3. In the lower energy isomer 2I, the additional solvent water is in the second solvation shell (S2). In isomer 2II, both solvent waters are in the first solvation shell and interact directly with the aqua ligand. The calculated energetic difference between these two isomers is quite small, with ΔE = 74 cm−1 and ΔG(185 K) = 156 cm−1, both favoring isomer 2I. Comparison of the experimental and calculated spectra indicates that both isomers are present under the experimental conditions. The lowest frequency peak at 2930 cm−1 and the shoulder at 3290 cm−1 are assigned to the LD and S1AD stretches of 2I, respectively. The dominant peak at 3355 cm−1 is assigned to the symmetric and antisymmetric OH stretches of the double H-bond-donor aqua ligand (LDD) of isomer 2II. The sharp features at 3633 and 3717 cm−1 correspond to the symmetric and antisymmetric OH stretches of the single acceptor water solvent, SA, in both isomers. With the addition of the third solvent molecule, the number of possible solvation geometries increases significantly. The three isomers that have the best agreement with the experimental spectrum are shown in Figure 4. Note that the isomers are labeled according to their relative ordering for ΔG(185 K); the five lowest energy isomers are shown in the Supporting Information. Isomer 3I has all solvent water molecules arranged in a “linear” H-bonded network and is an extension of structure 2I. Isomer 3III is the extension of 2II with the additional solvent water interacting with one of the waters in the first solvation shell. Isomer 3IV is the extension of 2I, with the additional solvent water binding to S1. These two isomers are calculated to have free energies of 266 and 362 cm−1 higher than 3I at 185 K. Comparison with the experimental spectrum indicates that all of the observed features can be attributed to isomer 3I. The very broad feature between 2600 and 3100 cm−1 is assigned to a strongly Hbonded LD feature of 3I. The features at 3170 and 3360 cm−1 match the two other H-bonded OH stretches (S1AD and S2AD) in 3I. The calculated vibrational transitions of isomers 3III and 3IV all overlap with those of 3I, and their presence cannot be ruled out. The three lowest energy isomers of [Ru(H2O)]2+·(H2O)4 are shown in Figure 5. Unlike the smaller clusters, the addition

together with three free OH peaks at 3633, 3660, and 3717 cm−1. The spectrum also displays a very broad and red-shifted feature spanning the range between 2600 and 3100 cm−1. The spectrum of [Ru(H2O)]2+·(H2O)4 is the most complex of the series, with many partially resolved features in the H-bonded region at 3195, 3275, 3370, and 3430 cm−1 as well as several partially resolved peaks in the free OH region at 3645, 3680, 3697, and 3730 cm−1. Figure 2 shows the calculated geometries and vibrational spectra of the D2-tagged [Ru(H2O)]2+·(D2)2 complex. To

Figure 2. Experimental (black) and calculated (red) IR spectrum of [Ru(H2O)]2+(D2)2 and [Ru(H2O)]2+(H2O)1. For ease of comparison, the calculated LD transition is convoluted with a 2σ = 40 cm−1 Gaussian while the other transitions are convoluted with a 2σ = 7 cm−1 Gaussian. Peak label: see Figure 3 caption, D2 = D2 stretch, CH = tpy/ bpy CH stretches, s/as = symmetric/antisymmetric H2O stretch.

simplify the geometry figures, the hydrogen atoms on the pyridine rings are omitted. The two weakly bound D2 are found to be located on the aqua ligand. The calculated vibrational spectrum of [Ru(H2O)]2+·(D2)2 is in excellent agreement with the experimental spectrum, which allows for easy assignment of the experimental spectral features. The peaks at 3654 and 3576 cm−1 are assigned to the antisymmetric and symmetric stretch vibrations of the aqua ligand, respectively. The calculated spectrum also includes the weak pyridine C−H stretches between 3050 and 3100 cm−1 and the stretch of the D2 tags35 at C

DOI: 10.1021/acs.jpca.5b04778 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Figure 3. Experimental (black) and calculated (red) IR spectra of [Ru(H2O)]2+(H2O)2. For ease of comparison, the calculated OH stretch below 3100 cm−1 is convoluted with a 2σ = 150 cm−1 Gaussian, the OH stretches at 3200−3600 cm−1 are convoluted with a 2σ = 40 cm−1 Gaussian, and all other transitions are convoluted with a 2σ = 7 cm−1 Gaussian. Peak label: The water molecules are labeled according to solvation layer (color coded and circled here), L = aqua ligand, S1 = first solvation layer, S2 = second solvation layer, etc. The subscript denotes H-bonding types: D = H-bond donating, A = H-bond accepting. LF = the free OH stretch of the aqua ligand.

Figure 4. Experimental (black) and calculated (red) IR spectra of [Ru(H2O)]2+(H2O)3. For ease of comparison, the calculated OH stretch below 3000 cm−1 is convoluted with a 2σ = 200 cm−1 Gaussian, the OH stretches at 3100−3600 cm−1 are convoluted with a 2σ = 40 cm−1 Gaussian, and all other transitions are convoluted with a 2σ = 7 cm−1 Gaussian. Peak label: see Figure 3 caption.

observed in the infrared spectra of H+(H2O)n clusters, which are also assigned to the SAAD and SAD vibrations.25

of the fourth solvent water produces an isomer that is distinctively more stable in terms of ΔG. Isomer 4I has a four-membered ring, with the last water (S3A) bound to the S2AAD water. The next isomer, 4II, has a five-membered water ring, while the third isomer, 4III, is the open equivalent of 4II with a symmetrically branched structure. These two isomers are calculated to have free energies of 627 and 774 cm−1 higher than 4I at 165 K, the experimental temperature. The congestion in the spectrum of [Ru(H2O)]2+·(H2O)4 makes it difficult to provide a detailed assignment and rule out the presence of multiple isomers. We note that all features in the experimental spectrum can be accounted for by the vibrational modes of isomers 4I and 4III. Two of the H-bonded features of 4II fall on the high-frequency edge of the broad cluster of peaks in the experimental spectrum, so there may also be minor contributions from this isomer. The two smaller peaks at 3645 and 3730 cm−1 are very close in frequency to the symmetric and antisymmetric stretches of free H2O and therefore can be assigned to the symmetric and antisymmetric stretches of SA waters in the outermost solvation shell. The partially resolved 3680 and 3697 cm−1 peaks likely correspond to the vibrations of SAAD and SAD waters, respectively. Interestingly, a pair of peaks at similar frequencies have been

■

DISCUSSION The assignment of the experimental spectra and the determination of the water network motifs allow us take a closer look at the effect of solvation on the [Ru(H2O)]2+ complex. We start with the bare [Ru(H2O)]2+ catalyst, whose spectrum displays only a pair of sharp free OH peaks. These are red shifted by 80−100 cm−1 compared to those of a free water molecule at 3756 and 3657 cm−1.41 The bare [Ru(H2O)]2+ calculation indicates that the presence of the D2 tags accounts for about one-half of the observed red shift, while the remaining ∼50 cm−1 of the red shift stems from the aqua ligand’s interaction with the ruthenium center. Such a red shift is characteristic of a water interacting with a charged center and comparable to those of other cationic metal−water clusters, such as M2+(H2O)6, where M = Mn−Zn.30 Upon the addition of the first solvent molecule, the H-bond-donating aqua ligand in [Ru(H2O)]2+(H2O)1 has an OH stretch frequency of 3200 cm−1. Further red shifting of the LD stretch, corresponding to increasing interaction strength in the aqua ligand’s H-bond, is observed in isomers exhibiting the linear chain solvent motif. D

DOI: 10.1021/acs.jpca.5b04778 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

are ∼400 cm−1 higher in frequency than the LD stretch of isomer 2I, indicating that the H-bond interaction is essentially split between the two hydrogen atoms of the aqua ligand. Similar trends can be observed in the LD/LDD and the S1 OH stretches of isomer 3III compared to 3I. Therefore, it appears that the stronger interactions in asymmetric solvation networks and their propagation into subsequent solvation shells lead to the stabilization of the unusual linear solvation motifs observed here. For most ions, electrostatic interactions between the charged center and the solvent molecules make the completion of each solvation shell more favorable than an extended linear network. Here, we observe clear evidence for the presence of such structures in [Ru(H2O)]2+·(H2O)2 and [Ru(H2O)]2+· (H2O)3. Electronic structure calculations confirm that these have favorable energetics and, as discussed below, provide some explanation for the prevalence of this solvation motif. The calculations provide information on the subtle changes in the geometry and electronic structure of the catalytic complex upon increasing solvation. Of particular interest is the Ru−OH2 bond length between the ruthenium center and the aqua ligand as a function of solvation, shown in Figure 6A. The

Figure 5. Experimental (black) and calculated (red) IR spectra of [Ru(H2O)]2+(H2O)4. For ease of comparison, the OH stretches at 3100−3600 cm−1 are convoluted with a 2σ = 40 cm−1 Gaussian and all other transitions are convoluted with a 2σ = 7 cm−1 Gaussian. Peak label: see Figure 3 caption. Figure 6. Calculated bond lengths of (A) Ru−OH2 (between ruthenium and aqua ligand oxygen) and (B) average of aqua ligand O−H as a function of the number of solvent molecules for all isomers depicted in Figures 2−5

Specifically, the LD stretch frequency is red shifted to 2930 cm−1 in 2I and to ∼2800 cm−1 in 3I. In typical M2+(H2O)n clusters, the H-bonded OH stretch of a first solvation shell water appears at ∼3400 cm−1 upon the formation of the second solvation shell.30 Therefore, it appears that the aqua ligand in the [Ru(H 2 O)] 2+ is involved in stronger H-bonding interactions than water in the first solvation shell of a bare metal dication. We note that the unusually low frequency of the aqua ligand’s OH stretch is more akin to those of the neutral water molecules in the first solvation shell of small H3O+(H2O)n clusters, which have frequencies of about 3000−3200 cm−1.25 Hence, this is direct evidence that the interactions between the solvent waters and the aqua ligand are stronger than typical water−water interactions and lay the foundation for proton transfer and formation of the [Ru− OH]+(H3O)+(H2O)n species. This strong H-bonding interaction persists beyond the first solvation shell, as indicated by the red-shifted OH features between 3100 and 3300 cm−1 of the S1 water of isomers 2I and 3I. In contrast, the structures with a more symmetrically solvated aqua ligand produce comparably weaker H-bond interactions. For example, the two LDD stretches of isomer 2II

Ru−OH2 bond length shortens with each additional solvent molecule, from 2.209 Å in [Ru(H2O)]2+ to about 2.15 Å for all three isomers of [Ru(H2O)]2+·(H2O)4. This indicates a general increase in strength of the Ru−OH2 interaction with increasing solvation. Figure 6B shows the calculated average O−H bond length of the aqua ligand, which goes from 0.965 Å in [Ru(H2O)]2+ to about 0.985 Å in [Ru(H2O)]2+·(H2O)4. Together, these trends indicate that the elongation of the aqua ligand’s O−H bond is compensated by a corresponding strengthening of the aqua ligand’s interaction with the metal center. To visualize how these changes are related to charge transfer, we constructed isosurfaces for the total SCF density differences between the entire [Ru(H2O)]2+·(H2O)n cluster ion and each of the individual [Ru]2+ and H2O subunits. These plots, presented in Figure 7, emphasize the change in electron density that occurs when the aqua ligand and solvating water molecules interact with the [Ru]2+ complex. Upon binding the E

DOI: 10.1021/acs.jpca.5b04778 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

continues to increase the electron density in the Ru−OH2 bond, yielding a yet shorter bond length. These electron density difference plots explain the stability of the linear chain solvation structure. Specifically, they show that the strong Hbonding in the asymmetric solvation networks propagates through successive solvation shells, yielding a highly polarized solvent water in the first solvation shell that interacts strongly with the aqua ligand. The stabilization of the symmetrical solvation motif does not become dominant until n = 4. In both cases, the solvation results in electron transfer to the Ru−OH2 bond, effectively strengthening it. We note that this effect, in which the H-bond network at the aqua ligand site influences the electronic structure of the complex, is consistent with the observation that the H-bond dynamics play a role in controlling the nonradiative decay of the excited states in this complex.42 The previously observed inverse kinetic isotope effect can therefore stem from the stronger H-bond interactions involved in D2O.

■

CONCLUSIONS The infrared predissociation spectra of [Ru(H2O)]2+·(H2O)n, n = 0−4, in the OH stretch region are reported and used to reveal the nature of the H-bonding network at the active site of this WOC catalyst. The analysis of the spectra and electronic structure calculations highlight the formation of strong Hbonds between the aqua ligand and the first solvation shell. These interactions are found to propagate through the subsequent solvation shells, which results in the stabilization of the linear solvation motifs for these microsolvated clusters. SCF difference plots show that these strong H bonds are promoted by the electron transfer from the H atom of the aqua ligand to the Ru−OH2 bond. This highlights the role of the solvent H-bond network in coupling the removal of the proton with the stabilization of the Ru−OH2 bond, leading toward the formation of [Ru−OH]+·(H3O)+(H2O)n. A more complete picture of the PCET in this system will require a similar study of the solvent network around the first reaction intermediate [Ru−OH]2+ to highlight the solvent reorganization involved in the next e−/H+ removal step. Such a study will involve the isolation of intermediate species in the catalytic cycle and is currently underway.

Figure 7. Electron density difference map (isovalue = 0.009) of selected isomers of [Ru(H2O)]2+(H2O)n, n = 0−3. These maps are constructed by subtracting the SCF density of the individual [Ru]2+ and (H2O) subunits from the entire [Ru(H2O)]2+(H2O)n complex. The terpyridine ring, oriented horizontally and out of plane, is omitted for clarity (only the nitrogen atoms are shown).

■

aqua ligand, the [Ru(H 2O)]2+ complex shows charge reorganization on the O and Ru atoms and an increase of electron density between them, establishing the Ru−OH2 bond. Interestingly, the plot also shows some increase in electron density on the nitrogen atom trans to the aqua ligand, demonstrating its role in stabilizing the complex. Upon addition of the first solvent water, there is a significant loss of electron density on the aqua ligand’s H atom involved in H-bonding. Most of this electron density moves to the Ru−OH2 bond, effectively strengthening it. The differences in the asymmetrically and symmetrically solvated complexes are revealed upon addition of the next solvent molecule. For isomer 2I, further electron density is removed from the aqua ligand’s H-bonded H atom, which creates a large polarization of the water in the first solvation shell. In comparison, isomer 2II has less electron density loss on each aqua H, but the total density transferred to the Ru−OH2 bond is slightly larger. For isomer 3I, the additional solvent water has a minimal effect on the electron density of the aqua ligand, and the Ru−OH2 bond of isomer 3I is similar to that of isomer 2I. On the other hand, isomer 3III

ASSOCIATED CONTENT

S Supporting Information *

Experimental and computational detail; further analysis of [Ru(H2O)]2+·(H2O)1 spectrum; tables of experimental and calculated vibrational frequencies; harmonic spectra calculated using ωB97XD; calculated geometry parameters; calculated ΔG. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.5b04778.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Science Foundation under grant number CHE-1454086. The computational F

DOI: 10.1021/acs.jpca.5b04778 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

(35) Marsh, B. M.; Zhou, J.; Garand, E. J. Phys. Chem. A 2014, 118, 2063. (36) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09; Gaussian, Inc.: Wallingford CT, 2009. (37) Iogansen, A. V. Spectrochim. Acta, Part A 1999, 55, 1585. (38) Rozenberg, M.; Shoham, G.; Reva, I.; Fausto, R. Phys. Chem. Chem. Phys. 2005, 7, 2376. (39) Rozenberg, M.; Loewenschuss, A.; Marcus, Y. Phys. Chem. Chem. Phys. 2000, 2, 2699. (40) Brites, V.; Lisy, J. M.; Gaigeot, M. P. J. Phys. Chem. A 2015, 119, 2468. (41) Shimanouchi, T. Molecular Vibrational Frequencies. In NIST Chemistry WebBook, NIST Standard Reference Database Number 69; Linstrom, P. J., Mallard, W. G., Eds.; National Institute of Standards and Technology: Gaithersburg, MD; http://webbook.nist.gov, retrieved Dec 2014. (42) Hewitt, J. T.; Concepcion, J. J.; Damrauer, N. H. J. Am. Chem. Soc. 2013, 135, 12500.

resources used in this work are supported by National Science Foundation Grant CHE-0840494.

■

REFERENCES

(1) Concepcion, J. J.; Jurss, J. W.; Brennaman, M. K.; Hoertz, P. G.; Patrocinio, A. O. T.; Murakami Iha, N. Y.; Templeton, J. L.; Meyer, T. J. Acc. Chem. Res. 2009, 42, 1954. (2) Liu, X.; Wang, F. Coord. Chem. Rev. 2012, 256, 1115. (3) Zong, R.; Thummel, R. P. J. Am. Chem. Soc. 2005, 127, 12802. (4) Wasylenko, D. J.; Ganesamoorthy, C.; Koivisto, B. D.; Henderson, M. A.; Berlinguette, C. P. Inorg. Chem. 2010, 49, 2202. (5) Concepcion, J. J.; Tsai, M.-K.; Muckerman, J. T.; Meyer, T. J. J. Am. Chem. Soc. 2010, 132, 1545. (6) Wang, L.-P.; Wu, Q.; Van Voorhis, T. Inorg. Chem. 2010, 49, 4543. (7) McDaniel, N. D.; Coughlin, F. J.; Tinker, L. L.; Bernhard, S. J. Am. Chem. Soc. 2007, 130, 210. (8) Concepcion, J. J.; Jurss, J. W.; Templeton, J. L.; Meyer, T. J. J. Am. Chem. Soc. 2008, 130, 16462. (9) Tseng, H.-W.; Zong, R.; Muckerman, J. T.; Thummel, R. Inorg. Chem. 2008, 47, 11763. (10) Masaoka, S.; Sakai, K. Chem. Lett. 2009, 38, 182. (11) López, I.; Ertem, M. Z.; Maji, S.; Benet-Buchholz, J.; Keidel, A.; Kuhlmann, U.; Hildebrandt, P.; Cramer, C. J.; Batista, V. S.; Llobet, A. Angew. Chem., Int. Ed. 2014, 53, 205. (12) Angeles-Boza, A. M.; Ertem, M. Z.; Sarma, R.; Ibanez, C. H.; Maji, S.; Llobet, A.; Cramer, C. J.; Roth, J. P. Chem. Sci. 2014, 5, 1141. (13) Huynh, M. H. V.; Meyer, T. J. Chem. Rev. 2007, 107, 5004. (14) Meyer, T. J.; Huynh, M. H. V. Inorg. Chem. 2003, 42, 8140. (15) Cukier, R. I.; Nocera, D. G. Annu. Rev. Phys. Chem. 1998, 49, 337. (16) Hammes-Schiffer, S.; Stuchebrukhov, A. A. Chem. Rev. 2010, 110, 6939. (17) Wasylenko, D. J.; Ganesamoorthy, C.; Henderson, M. A.; Koivisto, B. D.; Osthoff, H. D.; Berlinguette, C. P. J. Am. Chem. Soc. 2010, 132, 16094. (18) Chen, Z.; Concepcion, J. J.; Hu, X.; Yang, W.; Hoertz, P. G.; Meyer, T. J. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 7225. (19) Concepcion, J. J.; Jurss, J. W.; Norris, M. R.; Chen, Z.; Templeton, J. L.; Meyer, T. J. Inorg. Chem. 2010, 49, 1277. (20) Polyansky, D. E.; Muckerman, J. T.; Rochford, J.; Zong, R.; Thummel, R. P.; Fujita, E. J. Am. Chem. Soc. 2011, 133, 14649. (21) Tong, L.; Inge, A. K.; Duan, L.; Wang, L.; Zou, X.; Sun, L. Inorg. Chem. 2013, 52, 2505. (22) Angeles-Boza, A. M.; Roth, J. P. Inorg. Chem. 2012, 51, 4722. (23) Asmis, K. R.; Neumark, D. M. Acc. Chem. Res. 2011, 45, 43. (24) Bandyopadhyay, B.; Reishus, K. N.; Duncan, M. A. J. Phys. Chem. A 2013, 117, 7794. (25) Shin, J.-W.; Hammer, N. I.; Diken, E. G.; Johnson, M. A.; Walters, R. S.; Jaeger, T. D.; Duncan, M. A.; Christie, R. A.; Jordan, K. D. Science 2004, 304, 1137. (26) Johnson, C. J.; Dzugan, L. C.; Wolk, A. B.; Leavitt, C. M.; Fournier, J. A.; McCoy, A. B.; Johnson, M. A. J. Phys. Chem. A 2014, 118, 7590. (27) Miller, D. J.; Lisy, J. M. J. Am. Chem. Soc. 2008, 130, 15381. (28) Ke, H.; van der Linde, C.; Lisy, J. M. J. Phys. Chem. A 2014, 118, 1363. (29) Ohashi, K.; Sasaki, J.; Yamamoto, G.; Judai, K.; Nishi, N.; Sekiya, H. J. Chem. Phys. 2014, 141, 214307. (30) O’Brien, J. T.; Williams, E. R. J. Phys. Chem. A 2011, 115, 14612. (31) Bush, M. F.; O’Brien, J. T.; Prell, J. S.; Wu, C.-C.; Saykally, R. J.; Williams, E. R. J. Am. Chem. Soc. 2009, 131, 13270. (32) Cooper, T. E.; O’Brien, J. T.; Williams, E. R.; Armentrout, P. B. J. Phys. Chem. A 2010, 114, 12646. (33) Takeuchi, K. J.; Thompson, M. S.; Pipes, D. W.; Meyer, T. J. Inorg. Chem. 1984, 23, 1845. (34) Kamrath, M. Z.; Relph, R. A.; Guasco, T. L.; Leavitt, C. M.; Johnson, M. A. Int. J. Mass Spectrom. 2011, 300, 91. G

DOI: 10.1021/acs.jpca.5b04778 J. Phys. Chem. A XXXX, XXX, XXX−XXX